PROJECT SUMMARY/ABSTRACT Compulsive behaviors are prominent, disabling, and often treatment-resistant symptoms of several neuropsychiatric disorders, including obsessive compulsive disorder (OCD). Dysfunction within fronto- subcortical brain structures is thought to underlie compulsive behaviors, but the precise circuits involved remain unknown. Currently, chronic (> 4 weeks) treatment with serotonin reuptake inhibitors (SRIs) provides the only effective pharmacological monotherapy for compulsive behaviors; yet, approximately 50% of OCD patients do not respond to SRIs. We have shown that serotonin 1B receptors (5-HT1BRs) regulate the expression of compulsive behaviors. 5-HT1BRs are located on axon terminals of serotonin-containing neurons (presynaptic), and neurons containing other neurotransmitters (postsynaptic)(3), where they inhibit neurotransmitter release when activated. Furthermore, 5-HT1BRs signal through both a canonical G protein- mediated pathway, and a noncanonical G protein-independent pathway. Canonical 5-HT1BR-mediated Gi- signaling requires direct interaction of 5-HT1BRs with glycogen synthase kinase-3 beta (GSK3?). On the other hand, the intracellular scaffolding protein beta arrestin-2 (?-arrestin2) mediates noncanonical 5-HT1BR signaling. We recently found that activation of 5-HT1BRs within the orbitofrontal cortex (OFC) is necessary and sufficient to induce compulsive behaviors in mice. The OFC contains both pre- and postsynaptic 5-HT1BRs, including those on dorsal raphe-OFC and basolateral amygdala (BLA)-OFC projections, respectively. We propose to develop and use a two virus, in vivo CRISPR-Cas9 system to dissect the role of 5-HT1BR density and canonical versus noncanonical signaling within these two circuits in modulating compulsive behaviors. In Specific Aim 1, we will either overexpress or knockout 5-HT1BR expression within these two projections. We will infuse 1) a Cre recombinase (Cre) dependent adeno-associated virus (AAV) which expresses either 5- HT1BR, or the S. aureus Cas9 (SaCas9) gene plus a guide sequence against 5-HT1BR, into the dorsal raphe or BLA, and 2) a Cre-expressing retrograde canine adenovirus (CAV2-cre) into the OFC. In Specific Aim 2, we will also use the two virus, in vivo CRISPR-Cas9 system, but will infuse a Cre-dependent AAV expressing SaCas9 and a guide sequence against either ?-arrestin2 or GSK3? into the dorsal raphe or BLA. Mice will be evaluated for 5-HT1BR agonist-induced compulsive behaviors in the open field, a delayed alternation task, and an operant paradigm assessing both acquisition and persistence of habitual lever pressing. The proposed work could establish a novel in vivo CRISPR-Cas9 system for manipulating gene expression within specific neural circuits, and could lead to innovative therapeutic strategies for treating compulsions.